The Application Of Statistical Methods In Agricultural Environmental Monitoring And Evaluation

Agriculture is the most important basic industry in Chinese country economy. In the course of the development of society and economy some issues on agricultural environment emerged, such as heavy metal pollution in soil, pollution of water resource and so on, which restricted the sustainable development of agriculture.The agricultural environmental monitoring is a basic work of agricultural environmental protection in our country and an important part of agricultural environmental management system. In the actual monitoring work, we often use sampling survey to obtain a representative sample. In specific statistical works, employing what investigation methods, the amount of sample units, experimental cost, experimental precision and analysis of experimental data are all our common concerns. This study took the establishment of multi-stage sampling optimum test plans on rice cadmium detection for Jiangsu province, and statistical analysis of fishery water quality monitoring data of major lakes in Jiangsu province as the two cases, in order to discuss the application of statistical methods in the agricultural environmental monitoring and evaluation. Firstly, on the sampling strategy issue in statistical experimental methods, we discussed multi-stage sampling optimum test plans on rice cadmium detection for Jiangsu province. Secondly, in terms of main lakes fishery water quality monitoring in Jiangsu province, nested design variance analysis of fishery water quality data of the five lakes during 10 years from 2001 to 2011 was performed for the first time, and TOPSIS method was employed to give a comprehensive comparison of water quality of the five lakes. On this basis matter element analysis was used to optimize the layout of water quality monitoring points of Taihu Lake, Ge Lake, Hung-tse Lake, GaoBaoShaoBao Lake, and Luoma Lake. Finally, the lake water quality comprehensive evaluation system was built based on the past fishery water quality data of the five lakes in Jiangsu province. The main results are as follows:(1)A three stage sampling survey was taken as a real example to investigate the optimal plan of the sampling survey for cadmium in rice in Jiangsu Province. F-test showed that differences of the survey among cities and sampling villages were both very significant, and the error variance among cities was the largest in the experiment. In this paper, considering the survey expenses, the optimal sampling plan was explored under certain accuracy. The results indicated that twenty villages or agricultural extension stations could be selected with twice repeats at random from six randomly sampled cities to increase the precision of the survey. If the finance is limited, ten villages or agricultural extension stations from each city can be sampled alternatively to ensure the current sampling accuracy.The results provide an important reference for similar sampling survey.(2)For the first time nested design variance analysis of fishery water quality data of the five lakes (GaoBaoShaoBo Lake, Luoma Lake, Hung-tse Lake, Taihu Lake and Ge Lake)during 10 years from 2001 to 2011 (except 2004) was performed in order to compare the difference of the monitoring indicators among the five above-mentioned lakes in Jiangsu province. And TOPSIS method was employed to give a comprehensive comparison of water quality of the five lakes. The results indicated that the difference of 14 major water quality indicators was very significant among lakes except copper. In addition, transparency, total nitrogen and total phosphorus had very significant difference among stations for each lake; pH, chemical oxygen demand, oil, total phosphorus, lead, cadmium and mercury had significant or very significant difference among years for each station. The TOPSIS results showed that the fishery water quality of GaoBaoShaoBo Lake was the best, and Luoma Lake was just second to it, followed by Hung-tse Lake, Taihu Lake and Ge Lake. In combination with the geographic position of each lake, it showed that fishery water quality of the five investigated lakes was basically increasingly better from the south to the north in Jiangsu province, and the tendency revealed high association with the developed industrial economy. Statistical analysis of the data, which broke the limitations of statistical inference in the past for such data, can provide important reference for statistical analysis of the similar data.(3) Matter element analysis is an effective method to manage incompatible problems. This article based on water quality long-term localization monitoring data (DO, CODmn, TN, NH3-N, and TP) of main lakes of Jiangsu province from 2001 to 2011 (except 2004), and matter element analysis was used to optimize the layout of water quality monitoring points in the main lakes. The results indicated that the matter element analysis optimized the original 21 monitoring points to 15 points in Hung-tse Lake,13 points to 10 points in Ge Lake,18 points to 14 points in GaoBaoShaoBo Lake,14 points to 11 points in Luoma Llake, and 35 points to 27 points in Taihu Lake, respectively. These results would be helpful to improve monitoring efficiency.(4) The lake water quality comprehensive evaluation system was built on the basis of previous major lakes fishery water quality monitoring data in Jiangsu province. The system can achieve the bulk import of Jiangsu five lakes fishery water quality monitoring data, data display, the increase or decrease of the monitoring stations and other functions, which is a special computer software for lake fisheries water quality analysis.